Nucleotide Binding to the Hydrophilic C-terminal Domain of the

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1994 by The American Society for. Biochemistry and MoleculaI’ Biology, Inc.

Val. 269,No. 19,Issue of May 13,pp. 14032-14037, 1994 Printed in U.S.A.

Nucleotide Binding to the Hydrophilic C-terminal Domain of the Transporter Associated with Antigen Processing (TAP)* (Received for publication, November 17, 1993, and in revised form, January 27, 1994)

Kristian M. Muller+§,Christoph EbenspergerSO, and Robert TampBSfll From the $Max-Planck-Znstitut fur Biochemie, 0-82152 Martinsried, Germany and the Wehrstuhl fur Biophysik E22, Technische Universitat Munchen, 0-85748 Garching, Germany

The gene products of tapl and tap2 encoded in the Studies of mutant cell lines of human (LBL721.174, T2) and major histocompatibility complex (MHC) class I1 region murine (RMA-S) origin as well as the discovery of allelic variabelong to the ATP binding cassette superfamily of trans- tions in the rat cim locus led t o the discovery of essential genes porters. They are thought to form a heterodimer for the for correct MHC class I antigen presentation in the MHC class delivery of peptides into the lumen of the endoplasmic I1 region. Subsequently, human t u p l (formerly named ring4 or reticulum; peptides are required for correct assembly p s f l ) (3, 4) and tup2 (ring11 or p s f 2 ) (5, 6) as well as the rat and presentation of the MHC class I molecule peptide t u p l and t u p 2 ( m t p l (71, m t p 2 (811, and mouse t a p l and tap2 complex at the cell surface. To elucidate the ATP binding ( h u m 1 (91, hum2 (10)) were identified and sequenced. properties of these proteins in vitro, we expressed the The amino acid sequence comparison of the TAP proteins hydrophilic C-terminalpart of human transporter asso- revealed high homology to theABC superfamily of transporters ciated with antigen processing (TAP1)(nucleotide bind- (11)also named traffic ATPases (12). In the human TAPl and ing domain (NBD)-TAPl, amino acids 452-748) and TAP2 TAP2, as well as in other species, the ATP bindingmotifs (NBD-TAP2, amino acids 399-686)fused to a His, tag in Walker A and Walker B (13) were identified. Hydrophobicity Escherichiacoli. Therecombinant proteins accumuplots of the human TAPl and TAP2 point to six to eight translated exclusively in inclusion bodies and were solubimembrane-spanning regions. lized under denaturing conditions. Afterpurification by Restoring correct antigen presentation by gene transfer to immobilizedmetalionaffinitychromatography, we were able to refold the domains for functional studies. defective mutants proved the involvement of human (14-161, NBD-TAP1 bound to C-8-ATP-agarose and was specifi- mouse (17), andrat (18)TAP in antigenic peptide presentation. cally eluted with ATP or EDTA. Photoaffinity labelingof These experiments also led to the conclusion that the transporter functions as a heterodimer, analogous to other ABC NBD-TAP1 with the ATP analogues 8-azid0-[y-~~P]ATP transporters. By immunogold labeling, TAPl was located in the and 3’-0-[(4-azido-3,5-[‘26I]diiodo-2-hydroxybenz~yl)-~alanyl]-ATP was specific. Theaddition of 50 p~ ATP in- endoplasmic reticulum membrane facing the hydrophilic Chibited photoaffinity labeling by 8-azido-ATP down to terminal partof the cytosol (19).While this paper was in prepa8% of controls. Efficiencyof inhibition decreased as fol- ration, further experiments supportedthis view; using permelows: ATP > GTP > ADP > CTP > AMP. Photolabeling of abilized cells the transport of peptides was found to be TAPNBD-TAP2 was not observed. ATP hydrolysis byNBDand ATP-dependent (20, 21). In addition, a microsomal assay TAPl was not detected. Untilnow strong but only indi- has been described demonstrating the requirement of TAPl rect data of the TAP function existed. The described exand ATP for peptide translocation into microsomes (22). periments demonstrate ATP binding to an isolated In contrast, there are reported experiments questioning the domain of the antigenic peptide transporter, TAP, and emerging model of ATP-dependent peptide translocation. As a therefore support the theory of ATP-dependent peptide possible explanation for the observation of peptide movement translocation. into microsomal vesicles in theabsence of ATP, a distinct function of the TAP transporter was mentioned (23). Peptide translocation into microsomes of defect and wild-type cells in the Immune response toendogenous proteins requires their deg- absence of ATP and after proteinase treatment (24) indicated radation, transport of the fragments into theendoplasmic re- that there isno need for ATP or even a transporter. ticulum lumen, their association with the assembling MHC’ In the diverse family ofABC transporters the nucleotide class I molecule, and finally, the presentation of the peptide- binding characteristics and ATPase activities are of general MHC complex on the cell surface (for review see Refs. 1and 2). interest. The ATP binding and hydrolysis properties of some members have been studied in more detail. These include the * This work was supported by Grant Ta157/2-1 from Deutsche For- eukaryotic cystic fibrosis transmembrane conductance regulashungsgemeinschaft. The costs of publication of this article were de- tor (CFTR) (25-271, the P-glycoprotein (multidrug-resistancetherefrayed in part by the payment of page charges. This article must fore he hereby marked “aduertisenent” in accordance with 18 U.S.C. associated protein) (28-33), and the yeasta-factor transporter STE6 (34) as well as the prokaryotic proteins MalK (35-37), Section 1734 solely to indicate this fact. 8 K. M. M. and C. E. contributed equally to this work. HisP (38, 39), and HlyB (40). The functional homology of the /I To whom correspondenceshould be addressed.Tel.: 49-89-8578- nucleotide binding domainsof the ABC transporters is demon2646;Fax: 49-89-8578-2641. The abbreviations used are: MHC, major histocompatibility com- strated through thechimeric proteins STEG-NBD1-CFTR (41) and HisP MalK (42). Based on the homology of ATP binding plex; TAP, transporter associated with antigen processing; NBD, nucleotide binding domain; ABC, ATP binding cassette; IPTG, isopropyl- proteins andbiochemical data, structuralmodels for the nuclegel electrophoresis; otide binding domains of ABC transporters have been postup-o-thiogalactoside; PAGE, polyacrylamide [‘251]ASA-pAla-ATP,3’-0-[(4-azido-3,5-[’25I1diiodo-2-hydroxybenzoy1)-~lated (43, 44). alanyl]-ATP; IMAC, immobilized metal ion affinity chromatography; In this paper we report the recombinant expression, purifiNTA, nitrilotriacetic acid; CFTR, cystic fibrosis transmembrane concation, and refolding of the hydrophilic domain of human TAPl ductance regulator.

14032

Nucleotide Binding to NBD-TAP1

14033

(50mM TrisiHCI, 150 mM concentration of 20 pg/ml with refolding buffer NaCI, 5 mM MgCI,, pH 7.3) and supplemented with 4M urea and 2 mM cysteine to prevent aggregation. The protein solution was dialyzed in three stepsof 12 h each: first, with refolding buffer supplemented with 2 M urea and 2mM cysteine; second, with refolding buffer supplemented with 2 mM cysteine; and third, with refolding buffer. All steps were MATERIALSAND METHODS performed at 4 "C. Standard Techniques-The protein concentration was determinedby DNA Constructs-The full-length cDNAof TAPlinthe vector pCDM-8 and TAP2 in thevector pBS was the kind giftof Dr. J. Trows- a bicinchoninic acid assay (Pierce Chemical Co.) or by measuring the dale (3). The hydrophilic C-terminal domain of TAPl andTAP2, includ- absorbance at 280 nm. Proteins were separated in gradientpolyacrylbuffer (2 mM dithiothreitol) ing the nucleotide binding motifs Walker A and B, Walker were selected amide gels(8-16%) using a reducing sample according to hydrophilicity plots; for binding studies to substrates or with additional denaturationfor 2 min at 95 "C and stained with Coomassie Brilliant Blue R or silver. Two-dimensional gel electrophoresis cofactors, as well a s for the generation of specific antibodies, the sefocusing on Immobiline dry strips (Pharlected sequence was not restricted to the conserved part of the domain. was performed with isoelectric macia LKB Biotechnology Inc., pH 5.0-8.0) in the first dimension and The C-terminal partof TAPl from amino acids 452-748 (counted from 12% SDS-PAGE in the second dimension. NBD-TAP1 was identified the second ATG (3)) was amplified by polymerase chain reactiondefined by theprimers 5'-CGTCTAGGATCCATCGAGGGTAGAAGTCAGTT- with the monoclonal antibody mAb148-3 raised against a C-terminal standard immunoblottechniques.'Microsequencing CACCCAGGCT-3' and 5'-ATGCCGGATCCTCATTCTGGAGCATCTGC- peptideusing amino acid analysis was performed as described (46). 3'. The expressed domain of TAP2 spanning amino acids 399-686 was Nucleotide Affinity Chromatography-To testnucleotidebinding, defined by theprimers 5"CGGCTTGGATCCATCGAGGGAGAATGCAGGCTGGGGAGCTC-3' and 5'-TAGCCGGATCCTCAGTCCATCAG- IMAC-purified and refolded domains were applied to blue 2-Sepharose (Pharmacia) and C8-ATP-agarose (Sigma). Reactive dye chromatograCCGCTG-3'. The primers introduced a BamHI restriction site. In addition, the forward primer inserted a Factor Xa cleavage site, not used phy was performed in a buffer containing 50 mM TrisiHCI, 20% (vh) glycerol, 5 mM MgCl,, and 2mM dithiothreitol, pH7.5. For ATP-agarose in the experiments described here. Polymerase chain reaction was performed accordingt o standard protocols (1min at 94"C, denaturation; 1 chromatography, refolding buffer (50 mM Tris, 150 mM NaCI, and 5 mM min a t 40 "C, annealing; 3 min a t 72 "C, elongation). The polymerase MgCl,, pH 7.3) was used. For washing or elution, these buffers were supplemented with 1 M NaCl or 1 mM Na$TP, respectively. Alternachain reaction products were digested with BamHI and gel purified (Geneclean, Bio 101, Federal Republic of Germany). The generated endstively, a buffer containing 20mM EDTA instead of MgCI, was used for elution from the ATP-agarose. were compatible with in-frame insertion into the BamHI site of the ATP Photoaffinity Labeling-Photoaffinity labeling of IMAC-puripQE-9 expressionvector (Diagen, FRG). The vector was linearized with BamHI, dephosphorylated with calf intestinal phosphatase, and ligated fied and refolded domains was performed with two different ATP derivatives: 8-a~ido-[y-~'P]ATP (ICN) withspecific a activity of 314 GB¶/ with the inserts to generate constructscoding for the N-terminal His, mmol and 3'-0-[(4-azido-3,5-['25IJdiiodo-2-hydroxybenzoyl)-~-alanylltag NBD-TAP1 (amino acids452-748) or -TAP2 (amino acids399-686); with a specific activity of approximately 600 the proteins were namedNBD-TAP1 and NBD-TAP2. The ligated plas- ATP ([12511ASA-pAla-ATP) by the chloramine-T mids were transfected into Escherichiacoli MI5 (Diagen) carrying the GBq/mmol. Iodination of the latter was performed method just before use. The photoreactive group wascoupled to the ripREP4repressorplasmid.Correctrecombinantswereidentified by bose 3'-OH group via a spacer molecule. This photolabel was kindly multiple restriction digests and complete sequencingof the inserts by .~ the dideoxynucleotide method (U. S. Biochemical Corp. Sequenase kit) provided by Dr. Edmund Bauerlein and Dr. Rongbao Z h a ~Reactions (45). For selection of the hydrophilic domains, amino acid sequences were performed in 500 pl of refolding buffer (50 mM Tris/HCl, 150 mM were analyzed with the GCG program package (University of Wiscon- NaCI, 5 mM MgCl,, pH 7.3) containing 0.3 PM (10 pg/ml) refolded protein and 0.48 p~ 8-azido-[y3'P1ATP or 0.3 p~ [1Z511ASA-pAla-ATP. The sin) using "Pepplot." Isoelectric points were calculatedby the program mixture was incubated on ice for 15 min and illuminated with a UV "Isoelectric." lamp (366 nm, 125 watts, 4 min, distance of 4 cm, 4 "C). To test bindProtein Expression a n d Purification-E. coli strains were grown a t ing specificity, Na&TP was added before and after the photoreactive 37 "C in LB medium, supplemented with 0.1 mg/ml ampicillin and 0.025 mg/ml kanamycin. Recombinant protein expression was induced ATP derivative. For binding studies, various competitors (AMP, ADP, a t a cell density correspondingt o an A,,, of 0.8 by the addition of IFTG ATP, CTP, GTP) and inhibitors (EDTA, N-ethylmaleimide) were incu(1mM final) and continued incubationfor 3 h. Unless otherwise stated, bated with the protein for 15 min before addition of the photoreactive the following procedures were performed a t 4 "C with a cell culture ATP derivative. Reactions were precipitated by the addition of trichloroacetic acid, and the proteins were analyzed by SDS-PAGE. The gels volume of 1 liter. After harvesting and washing in50 mM TridHC1, pH were stained with Coomassie and subjected to autoradiography (Amer8.0, the cells were incubated for 15 min a t room temperature in20 ml of lysis bufferI (100 mM Tris/HCI, 500 mM NaCI, 250 mgfliter lysozyme, sham, p-max, UK). Bands in gels and films were scanned and quanti40 mgfliter DNase I, 1mM phenylmethylsulfonyl fluoride, pH8.0).The fied with the program Image (NIH). Inhibition of labeling was calcucell suspension was disrupted with a Branson sonifier in a rosette vessellated according to film density/gel density divided by the value of the (12.7 mm tip, duty cycle 30%, output control 6,15 rnin). Protein aggre- uncompeted reaction (control). gates, membrane fraction, andcell debris were collected by centrifugaRESULTS tion a t 7,800 x g for 30 min. The supernatant is referred t o as the soluble fraction. The pellet was suspended in 40 ml of lysis buffer I1 (100 mM Expression and Purification of Nucleotide Binding DoTris/HCl, 500 mM NaCI, 0.5% (w/v) Triton X-100, pH 8.0). Inclusion mains-The hydrophilic C-terminal domains of human TAPl bodies were pelleteda t 7,800 x g for 30 min. This supernatant is defined (amino acid 452-748) and TAP2 (amino acid 399-686) were as the membrane fraction. The inclusion bodies were suspended in 2.5 cloned into a BamHI site of a pQE-9 expression vector for ml of freshly prepared solubilization buffer (25 mM Na,HPO,/NaH,PO,, 25 mM Tris/HCI, 500 mM NaCl, 8 M urea, 30 mM p-mercaptoethanol, pH synthesis of recombinant fusion proteins with an N-terminal 8.0).After solubilizationat room temperature for 30 min the suspension His, tag. Transfection and reproduction of the recombinant was diluted 1 : l O with the same buffer containing 6 M urea, omitting vector in E. coli M15 without an additional repressorplasmid p-mercaptoethanol, and incubated overnight. Insoluble particles were yielded only religated vectors or contained the insert in the removed by centrifugation a t 31,000 x g for 40 min and filtration wrong direction. Attempts to use a different vector system through a 45-pm filter. The solubilized NBD-TAP wassubjected t o immobilized metal ion affinity chromatography(IMAC) on an Ni-NTA- (pET-16b) also failed. The presence of pREP4 plasmid coding agarose column (Diagen) a t room temperature. Thecolumn was washed for a lac repressor allowed the amplification of correctly ligated with 1 bed volume of solubilization buffer containing 6 M urea, 3 mM vector constructs. Induction of recombinant protein expression P-mercaptoethanol, 10 mM imidazole, pH 8.0. The protein was eluted with 1mM IPTG at 37 "C for 3 h yielded protein of the predicted with a n imidazole gradient reaching100 nm imidazole in bed 7 volumes size (Fig. 1,A and B ) . or using step-by-step decreasing pH (pH 8.0, 6.3, 5.7, and 4.5). For NBD-TAP1 two bands were expressed after induction Refolding Procedure-Diverse methods were comparedfor refolding and purified by Ni-NTAaffinity chromatography (Fig.M). The of recombinant NBD-TAP1 and NBD-TAP2 from solutionssupplemented with 6 M urea and 3 mM f3-mercaptoethanol. Simple dilution with refolding buffer precipitated the protein quantitatively. For funcM. Wesse and R. Tamp6, manuscript in preparation. tional studies Ni-NTA-purified fusion proteins were diluted to a final E. Bauerlein and R. Zhao, unpublished data.

and TAP2. Their ability to bind nucleotides specifically is characterized, providing the first directevidence for the ATP affinity of TAP1. These data support themodel of an active peptide translocation mechanism.

NBD-TAP1 Nucleotide to Binding

14034

A M I ”

1

2

3

4

5

6

1 2 3 4 5 6 7 8 9 1 0

92.5.

7

8

9

1

0

1

1

M ”

- 92.5

67.0.

. 67.0

45.0.

.45 0

29.0.

. 29.0

21 .o . 12.5 .

- 21 .o

--“

- 12.5

B

M 92.5

1

2

3

4

4

protein was eluted between 45 and 75 mv imidazolr I Fic. I A . /ones 9 and IO ) or a t pH 5.7 and 4.5 ( Fig. 2. Ianr .I 1. Cnntaminants remained in this chromatography. In t h r c a w of NRD29.0 TAPI, purification to homogeneity \vas possihlr aftc.r succrssful 21 .o refolding and ATP affinity chromatography (see text hclmv and Fig. 2, lanes 9-11 J. 12.5 The identity of the expressed proteins was confirmrd in srvera1 ways. The molecular mass of NRD-T.4PI w a s eatimatrd FK:.1. Expression and purification of NRD-TAP1 and NRDfrom SDS-PAGE to be 34.7 kDa, which is in good akTrhr.mrnt TAPS. A, Inncs 1-10, Coomassir-stained polyacrylamidr grl (8-16‘i 1 o f with the calculated value of 34.2 kDa (including thc ifis,: tar: the expression and purification strps o f NRD-TAP1 using Ni-NTA afand Factor Xa site). The second larger protein hand o f S E I ) finity chromatography (imidaznlr gradirnt); Innrs I 1 and 12, immunohlot with thr monoclonal antihody mAb14X-3 detected NBD-TAP1 r x TAP1 corresponds to a protein about 1.6 kDa largrr than thr clusivrly in inclusion hodirs; lnnr M , molrcularweightmarker. expected NRD-TAP1. By two-dimensional gtl rlrctrophorrsis Exprrssion: lnnc. 1, wholr crll lysate o f E. c d i M I 5 a f t w ovrrnight t h e isoelectric point of NBD-TAP1 was detrrminrd to hr pl = ~ wholr cell lysate aRer 3 h o f induction with I P T G ; lnnr culturr; l n n 2. 6.42, whichis in good agreement with thr calculatrd valur o f pI 3 , soluhlr fraction of the crll lysate; Innr 4.membrane fraction; lnnr 5 . solubilizrd inclusion bodies applied to M A C . Purification hy Ni-NTA = 6.45 (data not shown). The spot that shiftrd to :I higher affinity chromatokvaphy: lnnr 6. flow-through; Innrs 7 and 8.wash at 10 molecular mass was slightly mora acidic. with pI = 6.32. Sotc mal imidazole; Innr 9 , elution a t 15 m\l imitlazolr; lnnr I O , rlution a t 50 t h a t both bands were purified due to the S-trrminal His,, tag msl imidazole. Irnmunohlot: lnnr 11, snluhlr cell lysate and mrmhranr and identified with the monoclonal nntihody (mAhl48-.3 I raised 11. Coomassie-stainrd fraction; lnnr 12, soluhilizrdinclusionbodirs. polyacrylamidr gel t8-lfir; 1 o f thr rxprrssion and purification strps o f against the C terminus of TAPI. Furthrrmorr, S-trrminal srNRD-TAP2 using Ni-NTA affinity chromatographv( p H steps). 1,nrw M , quencing of the first three residues (Met-Arg-Glyr o f both bands molecular Wright markrr. Exprrssion: /ant, I. wholr crll lysate o f E . coli of S l 3 I ) confirmed identity with thr predicted start srqurncc M15 aftrr ovrrnight culture;Innc, 2, whole cell lysate aftrr 3 h of inducTAP1. In addition, both bands werr sprcifically Iahrlrd \c.ith tion with I P T G . I’urification hy Ni-NTA affinity chromatography: inn^ 3. flow-through, p€l 8.0; /nnr 4 . rlution a t pH 4.5. rxphotoreactiveATPderivatives. Thest. findingsmighthr plained by an overread eukaryotic stop codon; thc. rrsulting recombinant protein accumulated exclusively in inclusion hod- protein would he 1 kDa larger than the NRD-TAP1 tcrminatcd at the appropriate stop codon and have a calculatrd pI of 6.27. ies as determined by immunoblot of the soluble, membrane, In the caseof NBD-TAP:!, the expression and purificationby and inclusion body fraction (Fig.l A , 1anc.s 11 and 12 ). Attempts or doubleexpressionfailed,sincehoth Ni-NTA affinity chromatography were identical to S13D-TAPl toavoidaggregation and summarized briefly in Fig. In. The identity of SRI)-TA1’2 bands were also found in aggregates when the induction temperature was lowered to 26 “C and lower concentrations of was confirmed by measuring the molecular mass ( 3 5 kDn. talIPTG, or other methods to disrupt cells were used. The two-step culated 33.3 kDa) and the isoelectric point ( P I = 6.47. ralculatrd preparation of the inclusion bodies removed major impurities. PI = 6.51). NuclroticfrAffinity C h r o m n t o ~ r a p h ~ - I ~ C l A ~ - p u r and ifi~~~~ Solubilization of the inclusion bodies was achieved in two further steps starting with 8 JI urea and 30 m31 p-mercaptoetha- refolded NBD-TAP1 was suhjected to affinity chromatnfraphy on blue 2-Sepharose, known to mimic nucleotidrs. and on (’-% nol. After dilution of the denaturing agents to concentrations suitable for Ni-NTA affinity chromatography ( 6 51 urea, 3 mw linked ATP-agarose. NRD-TAP1 bound to hlur 2-Srpharosr and C-8-ATP-agarose. while contaminants were found in the flow/3-mercaptoethanol),thesuspensionwasfurtherincubated. Lower urea concentrations, lack of 0-mercaptoethanol, or solu- through. NRD-TAP1 was specifically rluted by 1 m v ATP from the blue Sepharose (data not shown I and CR-ATP-agarosr ( I’ic. bilization times under 2 h drastically reduced protein yields. 2, lanes 9-11 ) or 50 msl EDTA from the CR-ATP-agarosr. S131)The enriched recombinant protein was subjected to immohiTAP2 did not hind to either ATP-agarosc or hlur Scphnrnsr. lized Ni-NTA affinity chromatography, which was performed with either an increasing imidazole gradient from 10 to 100 mlr The specific hinding to the nucleotidr affinity matrix gavr first hints for a n ATP affinity of NRD-TAP1 and providcd :I mc.thod or step-by-stepdecreasingpH of 8.0,6.3, 5.7, and 4.5. The

-

Binding Nucleotide

to NBD-TAP1

for further purification, even from misfolded NBD-TAPl. Photoaffinity Labeling with Radioactive ATP DerivativesPhotoaffinity labeling with 8-a~ido-["~P]ATP has been used to study the ATP binding properties of several ABC transporters like CFTR (26-28), multidrug-resistant protein (29), STE6 (35), MalK (36, 37) and HisP (39-41).[1251]ASA-PAla-ATPis a new substanceinthisfieldwiththephotoreactiveazidosalicylic group attached to the ribose via the 0-alanyl spacer. First, labeling and specificity of labeling were examined. IMAC-purified and refolded NBD-TAP1 was labeled by 8-azido[y-""P]ATP and [12'1 IASA-PAla-ATP, respectively. The photolabeling of NBD-TAP1 by 8-a~ido-[y-'~PlATP and [1z51]ASA-pAlaATPwassignificantinbothcases.Evenafterverylong a exposure time, no additional bands were detectable although small amountof impurities is present. Without illumination no labeling occurred. The additionof 1 mM ATP before orafter t h e incubation with the photoreactive ATP derivative prevented labeling (data not shown). Labeling of NBD-TAP2 purified and refolded in the same way as NBD-TAP1 was not observed (data not shown). The influence of various substances on the 8-azido[y-:'2P1ATP labeling was examined (Fig. 3A). Fig. 3B summarizes the data obtained from two labeling experiments with independently IMAC-purified and refolded samples of NBDTAP1. Compared with the non-competing reaction, the labeling of NBD-TAP1 was decreased in the presence of the indicated substance as follows: 1 mM AMP, 64%; 1 mM ADP, 21%; 50 V M ATP, 8%; 1 mM ATP, 6%; 1 mM CTP, 55%; 1 mM GTP, 19?6; 1 mM N-ethylmaleimide, 83%; and 20mM EDTA, 41%). Labeling with [1251]ASA-0Ala-ATPwas reduced by increasing concentrations 11~4: M A T P67% , (lane 2); 110 ofATP and GTP as shown in Fig. VM ATP, 20% (lane 3 ) ; 1100 PM ATP, 15% (lane 4 ) ; 11 V M GTP, 43% (lane 5); 110 V M GTP, 26% (lane 6). Using a sensitive radioactive ATPase assay we were not able to detect ATPase activity of NBD-TAP1. The additionof 0.2 mM lauryldimethylamine N-oxide or 0.1 mM nonradioactive ATP did not stimulate the ATP hydrolyzation.

14035

A M 1

2

3

4

"

5

-

5

"

7

8

? ' r

C

21 .o.

DISCUSSION

The predicted functions of ATP binding and peptide translocation of the proteins derived from the tap1 and tap2 genes found in the MHC class I1 region have been questioned (23,24). FIG.3. Photoaffhity labeling of refolded NRD-TAP1 with To characterize the hydrophilic C-terminal domain of TAPl and 8-azido-[yqzP]ATP. The reaction mixture ( 5 0 m v Tri.dHCI.150 m v TAP2 and to directly address their ATP binding properties at a NaCI, 5 m y .MgCI,. pH 7.3, was supplemented with 0.3 p v h'i-NTAmolecular level, we expressed these domains in E. coli. Our purified and refolded NRD-TAP1 and 0.48 p~ FI-azido-[y-'TIATP. Competitors and inhihitors were added and incubated for 15A min. . laheling resultsregardingATPbindingcanbecomparedwithdata available from several other members of the superfamily of reactions were applied to SDS-PAGE. Upprr panel. Conmassir-stained gel; lower panel. corresponding autoradiography; luna M . molecular ABC transporters. The selected domains were not restricted to weight marker; lane 1 , no competitor;Ianr 2, with additional 1m\l AMP: the conserved region between the Walker A and Walker Bmo- lane 3, 1 my ADP; lane 4. 50 p~ ATP; Ianr .5, 1 msr ATP; l o n r 6 . 1 m\r CTP; lane 7. 1 mM GTP; lane 8. 1 r n M N-ethylmaleimidt-; I a n ~9,20 m v tifs, since ATP binding, as well as association with peptides, EDTA; lane 10, no competitor.R . inhibitory effectsof various suhstnnces other proteins, or cofactors may require additional structural on the aninity labeling of NBD-TAP1 are evaluated from the scanned elements. Furthermore, the protein was designed to allow the gel and autoradiography with the program image. Data were ohtainrd generation of specific antibodies. from the labeling experiment shown in Fig. 4.4 and a second experiment with independently purified and refolded NBI)-TAPl. In the case of Only a well controlled promoter system allowed the recombinant expressionof the hydrophilic C-terminal domainsof the CTP and N-ethylmaleimide ( N E M ) .larger differences between experiments are caused by saturation of t h e [&ray film in one experiment, human TAPl and TAP2. Toxicity of the expressed nucleotide binding domains would explain the problems that occurred fused N-terminal His, tag. More stringent washing with higher A without a repressor plasmid or other expression vectors.convincing reasonfor the presenceof two proteins of NBD-TAP1 is imidazole concentrations or lower pH did not improve purity. the overreading of the cloned eukaryotic stop codon and the Some proteins, NBD fragments or E. coli proteins, bound a s utilization of the following prokaryotic stopcodon. This view is tightly as the fusion protein. Proteins co-purified in this chroin agreement on the physical dataof the proteins. matography step, however, did not interfere with further exThe recombinant proteins were exclusively found in incluperiments analyzing nucleotide affinity. Thus, separation by sion bodies. Isolation and extensive solubilization of the inclu- nucleotide affinity chromatography was not necessary. First evidencefor ATP bindingof NRD-TAP1 was obtainedby sion bodies yielded highly enriched recombinant protein. Furthercontaminantswereremovedbyimmobilized Ni-NTA the specific binding to blue 2-Sepharose and C-8-ATP-agarose. Additionally these chromatographiesallow further purification affinity chromatography under denaturing conditions using the

Binding Nucleotide

14036 1

2

3

4

5

f

i

-

to NBD-TAP1

M

92 5

.450

.

loo

.290

t-7

1 O

O

50

0

c

(0 ATP;

F 100

150

" 1000 i 100

o GTP) [ p m o l / l ]

FIG.4. Photoaffinity lnbeling of refolded NBD-TAP1 with I'2D11ASA-PAla-ATP.The labeling reaction was performed in refolding burner (50 mhl Tris/HCl, 150 mh! NaCI, 5 m y MgCI,, pH 7.3) supplemented with Ni-NTA-purified and refolded 0.3 ph1 NBD-TAP1 and 3 phl I"'1IIASA-PAla-ATPdemonstrating the inhibition oflabeling by increasing concentrations of Na.fiTP and Na,GTP. A, the mixtures were applied to SDS-PAGE. Upprr pnnrl. Coomassie-stained gel; lower pnnrl. corresponding autoradiography; lane 1 , no competitor; lnnr 2, 11 p~ ATP; lnnr 3 , 110 ph! ATP; lanr 4 , 1100 ps1 ATP; lane 5 , 11 p~ GTP; lane 6 , 110 P M GTP; lane M , molecular weight marker. R , plot summarizing the competition of labeling a t various concentrations of ATP ( 0 )and GTP ( 0 )as evaluated by scanning of the gel and autoradiographic film.

even from misfolded NBD-TAP1. These experiments correlate with findings for the recombinant NBDI-CFTR (2.5)and MzIIK,~ which also bind to C-8-ATP-agarose. A purification procedure for MalK using blue 2-Sepharose or red agarose is described (36), whereas HisP did bind to blue agarose but wasrecovnot ered by the addition of ATP (38). Thephotoaffinitylabeling of NBD-TAP1with8-azido[y-""PIATP and ['2511ASA-PAla-ATPwas specifically inhibited by the addition of ATP, clearly demonstrating ATP affinity of the recombinant nucleotide binding domain. ['"I]ASA-PAlaATP, which is a new reagent for ATP affinity labeling with properties comparable with 8-azido-[y-"'PlATP despite the photoreactive azidosalicylic group, is bound to ribose via the p-al-

' E. Schneider, personal communication

any1 spacer. Becauseof the one-step radioiodination before use. which allows introduction of up to two iodine atoms, and its longerhalf-lifetimecomparedwith8-azido-l y-"PIATP, the handling and sensitivityof this reagent is good. ATP ( 8 4 labeling of control in the presence of 50 p t ATP, GC> in the presence of 1 m>r ATP) was the most potent inhihitor followed by 1 m31 GTP. 1 mxt ADP as well as 20 msr EDTA were less efficient but still good in competition. 1 mlr AMP (64r; I , 1 msj CTP (55%), and 1 mlr N-ethylmaleimide (84r4 I had a minor influence on photolabeling. Photoaffinity laheling experiments of NBD-CFTR(25),multidrug-resistantprotein(28).STE6 (34), MalK (36), HisP (38). and HlyB (401revealed ATP as a potent inhibitor with half-maximal inhibition in a range from 10-'to lo-" M. In most casesM$* was present but does not seem as a n to be essential (25, 28). When tested, GTP also served inhibitor, whereas ADP, AMP, and CTP exhibited equal or only minor effects (28,36,38). In general a quantitative comparison of these values is difficult due to different labeling conditions. Nevertheless there is conformity for decreasing inhihition in the order ATP > GTP > ADP > CTP. NBD-TAP2 was neither bound to ATP-agarose or blue Sepharose nor labeled by 8-azido-[y-:"PlATP or ['?"I IASA-pAla-ATP. As wild-type NBD-TAP1 and NBD-TAP2 amino acid sequences are 6 0 9 homologousandbothproteinsweretreated in the same way, these results are presently difficult to interpret. In this respect, incorrect folding of the recombinant protein is as likely as the possibility of a different function for the TAP2 protein. Although we were able to demonstrate ATP binding, no significant ATPase activity was detected. These results correlate with findings with a synthetic peptide of NRD-CFTR (261 and HisP (38). However, in this respect the nucleotide binding docytosolic mains of the ABC transporters behave differently. The as ABC domain of HlyB fused to glutathione S-transferase (401 well as MalK (36, 37) is active in hydrolyzing ATP. To summarize, we report the specific binding of nucleotides tothehydrophilicC-terminaldomain of TAPl.Comparison with otherABC transporters demonstrates homologous behavior and function, insofar as is possible, given the limited understanding of this group of active transporters. From this point of view TAPl is a typical member of the ARC transporter superfamily. A c k n o w l ~ ~ d ~ m r n ~ s -are W eindebted to Dr. Bauerlcin and Dr. Zhao for providing the ['"IIASA-pAla-ATP and the permission to publish the data, Dr. Lottspeichand K. Anderson for microsequcmcing. and Dr. Trowsdale for the gift of cDNA clones.

Nucleotide Binding to NBD-TAP1 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

29. 30. 31.

ska-Ziegler, B., Ziegler, A., Trawsdale, J., and Townsend, A. (1992) Nature 355,641-644 Attaya, M., Jameson, S., Martinez, C. K., Hermel, E., Aldrich, C., Forman, J., Fischer Lindahl, K., Bavan, M. J., and Monaco, J. J. (19921 Nature 355, 647449 Momburg, F., Ortiz-Navarrete, V., Neefjes, J.,Goulmy, E., van deWal, Y,Spits, H., Powis, S. J., Butcher, G. W., Howard, J. C., Walden, P., and Hammerling, G. J . (1992) Nature 360, 174-177 Kleijmeer, M., Kelly,A., Geuze, H.J., Slot, J . W., Townsend, A,, and Trowsdale, J. (19921 Nature 357,342-344 Neefjes, J.J.,Momburg, F., and Hammerling, G. J. (1993) Science 261, 769771 Androlewicz, M. J.,Anderson, K. S., and Cresswell, P. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 9130-9134 Shepherd, J. C., Schumacher, T.N. M., Ashton-Rickardt, P. G., Imaeda, S., Ploegh, H. L., Janeway, C.A., Jr., and Tonegawa, S. (1993) Cell 74,577-584 Koppelman, B., Zimmerman, D., Walter, P., and Brodsky, F. M. (1992) Proc. Natl. Acad. Sci. U. S. A . 89, 39083912 Levy, F., Gabathuler, R., Larsson, R., and Kvist, S. (1991) Cell 67,265-274 Hartman, J., Huang, Z., Rado, T.A,, Peng, S., Jilling, T., Muccio, D. D., and Sorscher, E. J. (1992) J. Biol. Chem. 267, 64554458 Thomas, P. J., Shenbagamurthi, P., Ysern, X., andPedersen, P. L. (1991) Science 251,555-557 Bear, C. E., Li, C., Kartner, N., Bridges, R. J., Jensen, T.J., Ramjeesingh, M., and Riordan, J. R. (1992) Cell 68,809-818 Cornwell, M. M., Tsuruo, T., Gottesman, M. M., and Pastan, I. (1987) FASEB J. 1,51-54 Shimabuku, A. M., Nishimoto, T., Ueda, K., and Komano, T.(1992) J. B i d . Chem. 267,43084311 Al-Shawi, M. K., and Senior, A. E. (1993) J. Biol. Chem. 268, 41974206 Ambudkar, S. V., Lelong, I. H., Zhang, J., Cardarelli, C. O., Gottesman, M. M.,

14037

and Pastan, I. (1992) Proc. Natl. Acad. Sci. U.S. A. 89,847243476 32. Sarkadi, B., Price, E. M., Boucher, R. C., Germann, U. A,, and Scarborough, G. A. (1992) J. Biol. Chem. 267,4854-4858 33. Hamada, H., and Tsuruo, T.(1988) J. Biol. Chem. 263, 1454-1458 34. Kuchler, K., Dohlman, H. G., and Thorner, J. (1993) J. Cell Biol. 120, 12031215 35. Dean, D. A,, Davidson, A. L., and Nikaido, H. (1989) Proc. Natl. Acad. Sci. U. S. A. 8 6 , 9134-9138 36. Walter, C., zu Betrup, K. H., and Schneider, E. (1992) J. Biol. Chem. 267, 8863-8869 37. Morbach, S., Tebbe, S., and Schneider, E. (1993) J. Biol. Chem. 268, 1861718621 38. Hobson,A. C., Weatherwax, R., andAmes, G. F.-L. (1984)Proc.Natl. Acad. Sci. U. S. A . 81, 7333-7337 39. Bishop, L., Agbayani, J . R., Ambudkar, S . V.,Maloney, P. C., and Ames,G. F:L. (1989) Proc. Natl. A c ~ Sci. . U. S. A. 86,6953-6957 40. Koronakis, V, Huges, C., and Koronakis, E. (1993) Mol. Microbiol. 8, 11631175 41. Teem, J.L., Berger, H. A,, Ostedgaard, L. S., Rich, D. P., Tsui, L.-C., and Welsh, M. J. (1993) Cell 73, 335-346 42. Schneider, E., and Walter, C. (1991) Mol. Microbiol. 5, 1375-1383 43. Hyde, S. C . , Emsley, P.,Hartshorn, M. J.,Mimmack, M. M., Gileadi, U., Pearce, S. R., Gallagher, M. P., Gill, D. R., Hubbard, R. E., and Higgins, C.F. (1990) Nature 346,362-365 44. Mimura, C. S., Holbrook, S. R., andAmes, G. F.-L. (1991jProc.Natl. Acad.Sci. U. S. A . 88, 84-88 45. Sanger, F., Nicklen, S., and Coulson, A. R. (1977)Proc. Natl. Acad.Sci. U. S. A . 74,54634467 46. Eckerskorn, C., Jungblut, P., Mewes, W., Klose, J., and Lottspeich, F. (1988) Electrophoresis 9, 830-838